In models of myeloid leukemogenesis developed in our laboratory we have recently been focusing on the role of tumor suppressors. Cyclin dependent kinase (cdk) inhibitors are tumor suppresses that play an important role in cell cycle control and have been inactivated in human tumors either by deletion, promoter hypermethylation or by mutation. In particular, in a high proportion of human AML, MDS and ALL the promoter region of the gene encoding the cyclin dependent kinase (cdk) inhibitor p15INK4b has been found to be hypermethylated. In the mouse, the gene encoding the cdk inhibitor p15INK4b (Cdkn2b) is located on Chr 4 proximal to Cdkn2a which encodes the cdk inhibitor p16INK4a. Additionally, Cdkn2a encodes, through alternative splicing, p19ARF, a positive regulator of the p53 tumor suppressor pathway. p15INK4b is particularly interesting because it is upregulated at the transcription level in myeloid cells by differentiation and growth-inhibiting cytokines and is highly expressed in mature cells of the monocytic lineage. In one area of research this year we have began examining myeloid leukemias from several murine model systems for alterations in INK4 gene transcription in an attempt to correlate inactivation of INK4a or b with specific oncogene-induced leukemias. The first are neoplasms induced by endogenous c-myb, activated by insertion of retrovirus into the genome, while the others are induced by a retrovirus carrying the c-myc gene. Our data show that most Myb related tumors do not express p15INK4b mRNA, whereas most Myc tumors do express this RNA. A lack of expression in the Myb tumors is not due to deletion or methylation. Further studies in vitro have indicated that the c-Myb itself can repress the expression of p15INK4b. In the case of the Myc tumors, Northern analysis revealed that mRNA and protein for p15INK4b were expressed in a majority of tumors whereas the mRNA for the p16INK4a gene, was rarely expressed. Interestingly, two Myc tumors had aberrant INK4 transcripts that were determined to be fusions of p15Ex1 with p16Ex2 or 3 and this was due to deletion of a region encompassing Ex1? for p19ARF. Therefore, in the case of inappropriate c-myc expression, our data would suggest that alterations in the Cdkn4a locus may be primarily due to attempts by the cell to inactivate p19ARF and therefore the p53 tumor suppressor pathway. Another area of research involves the generation of mice with loss of alleles for the genes encoding tumor suppressors so that we can analyze them for susceptibility to myeloid leukemia. We have prepared a knockout of p15INK4a and are comparing the homozygous, heterozygous and wild-type phenotypes for susceptibility to retrovirus-induced leukemia.